To test the hypothesis that increased local control may lead to improved long term survival, we have applied three-dimensional conformal radiation therapy (3DRTP) to escalate the dose to the tumor while keeping dose to critical organs below acceptable levels. Until recently, the focus of our development has been primarily on the geometrical shape of the field, that is, optimum coverage of the target and sparing of critical organs. Dose distribution within the field was adjusted by beam weight and simple modulators, i.e. blocks and wedges. To further improve the dose distribution and to escalate the dose, the use of intensity- modulated fields is needed. There are two components concerning the use of intensity-modulated fields: the optimization of intensity profiles and the delivery of such fields. A dose-based preliminary version of the optimization algorithm has been developed, with the intensity-modulated beams delivered by dynamic multileaf collimator (DMLC). Treatment of prostate cancer with this new method has begun. However, the present approach, although improved relative to our previous practice, has limitations which will be remedied by the research proposed in this project. Our specific goals are: (1) To further develop and enhance the optimization algorithms for intensity-modulated radiation therapy (IMRT). New features include dose-volume constraints, optimization over a pre-existing dose distribution, and different forms of objective functions. (2) To explore the potential benefits of non-coplanar beams using a combination of fast simulated annealing and the inverse methods. With this approach, we hope to identify class solutions for optimum beam number and directions which are disease-site specific. The class solutions will then be used as a starting points for optimization for individual patients. (3) To investigate the potential use of combined electron and photon treatment in IMRT. The characteristics of these modalities are complementary to each other: the finite range of electrons would spare the organs beyond the target, and the sharper penumbra of photons would protect organs adjacent to the target. Our optimization algorithm can simultaneously optimize beams of different modalities and energies, thus providing an efficient tool to explore such technique. (4) To study the effects of treatment uncertainties on IMRT and to device method to incorporate them in the optimization of intensity distribution and the design of DMLC Alternatively the uncertainties can be included in plan evaluation and their effects assessed.